def command(self):
required("", self.input_tumor)
required("", self.input_normal)
tmp_vcf = "{scratch}/{uuid}.vcf.gz".format(scratch=self.scratch,
uuid=uuid.uuid4())
# run vardict without removing non-somatic variants, and adding "SOMATIC" INFO field for somatic variants
vardict_cmd = "vardict-java " + required("-G ", self.reference_sequence) + \
optional("-f ", self.min_alt_frac) + \
required("-N ", self.tumorid) + \
optional("-r ", self.min_num_reads) + \
" -b \"{}|{}\" ".format(self.input_tumor, self.input_normal) + \
" -c 1 -S 2 -E 3 -g 4 -Q 10 " + required("", self.target_bed) + \
" | testsomatic.R " + \
" | var2vcf_paired.pl -P 0.9 -m 4.25 " + required("-f ", self.min_alt_frac) + \
" -N \"{}|{}\" ".format(self.tumorid, self.normalid) + \
" | " + fix_ambiguous_cl() + " | " + remove_dup_cl() + \
" | sed 's/Somatic;/Somatic;SOMATIC;/g' " + \
" | sed '/^#CHROM/i ##INFO=<ID=SOMATIC,Number=0,Type=Flag,Description=\"Somatic event\">' " + \
" | vcfstreamsort -w 1000 " + \
" | bcftools view --apply-filters .,PASS " + \
" | vcfsorter.pl {} /dev/stdin ".format(self.reference_dict) + \
" | bgzip > " + tmp_vcf + " && tabix -p vcf " + tmp_vcf
# annotate variants with dbSNP id
annotate_cmd = "bcftools annotate --annotation {} --columns ID ".format(self.dbsnp) + \
" --output-type z --output {} ".format(self.output) + tmp_vcf + \
" && tabix -p vcf {}".format(self.output)
# remove temporary vcf and tabix
rm_tmp_cmd = "rm " + tmp_vcf + "*"
return " && ".join([vardict_cmd, annotate_cmd, rm_tmp_cmd])
def command(self):
required("", self.input_tumor)
required("", self.input_normal)
freq_filter = (
" bcftools filter -e 'STATUS !~ \".*Somatic\"' 2> /dev/null "
"| %s -c 'from autoseq.util.bcbio import depth_freq_filter_input_stream; import sys; print depth_freq_filter_input_stream(sys.stdin, %s, \"%s\")' "
% (sys.executable, 0, 'bwa'))
somatic_filter = (
" sed 's/\\.*Somatic\\\"/Somatic/' " # changes \".*Somatic\" to Somatic
"| sed 's/REJECT,Description=\".*\">/REJECT,Description=\"Not Somatic via VarDict\">/' "
"| %s -c 'from autoseq.util.bcbio import call_somatic; import sys; print call_somatic(sys.stdin.read())' "
% sys.executable)
blacklist_filter = " | intersectBed -a . -b {} | ".format(
self.blacklist_bed)
cmd = "vardict-java " + required("-G ", self.reference_sequence) + \
optional("-f ", self.min_alt_frac) + \
required("-N ", self.tumorid) + \
optional("-r ", self.min_num_reads) + \
" -b \"{}|{}\" ".format(self.input_tumor, self.input_normal) + \
" -c 1 -S 2 -E 3 -g 4 -Q 10 " + required("", self.target_bed) + \
" | testsomatic.R " + \
" | var2vcf_paired.pl -P 0.9 -m 4.25 -M " + required("-f ", self.min_alt_frac) + \
" -N \"{}|{}\" ".format(self.tumorid, self.normalid) + \
" | " + freq_filter + " | " + somatic_filter + " | " + fix_ambiguous_cl() + " | " + remove_dup_cl() + \
" | vcfstreamsort -w 1000 " + \
" | " + vt_split_and_leftaln(self.reference_sequence) + \
" | bcftools view --apply-filters .,PASS " + \
" | vcfsorter.pl {} /dev/stdin ".format(self.reference_dict) + \
conditional(blacklist_filter, self.blacklist_bed) + \
" | bgzip > {output} && tabix -p vcf {output}".format(output=self.output)
return cmd
def command(self):
regions_file = "{scratch}/{uuid}.regions".format(scratch=self.scratch,
uuid=uuid.uuid4())
bed_to_regions_cmd = "cat {} | bed_to_regions.py > {}".format(
self.target_bed, regions_file)
call_somatic_cmd = " | {} -c 'from autoseq.util.bcbio import call_somatic; import sys; print call_somatic(sys.stdin.read())' ".format(
sys.executable)
freebayes_cmd = "freebayes-parallel {} {} ".format(regions_file, self.threads) + \
required("-f ", self.reference_sequence) + " --use-mapping-quality " + \
optional("--min-alternate-fraction ", self.min_alt_frac) + \
optional("--min-coverage ", self.min_coverage) + \
conditional(self.use_harmonic_indel_quals, "--harmonic-indel-quality") + \
optional("", self.params) + \
repeat(" ", self.input_bams) + \
"""| bcftools filter -i 'ALT="<*>" || QUAL > 5' """ + \
"| filter_erroneus_alt.py -V /dev/stdin " + \
conditional(self.somatic_only, call_somatic_cmd) + \
" | " + vt_split_and_leftaln(self.reference_sequence) + \
" | vcfuniq | bcftools view --apply-filters .,PASS " + \
" | bgzip > {output} && tabix -p vcf {output}".format(output=self.output)
# reason for 'vcfuniq': freebayes sometimes report duplicate variants that need to be uniqified.
rm_regions_cmd = "rm {}".format(regions_file)
return " && ".join([bed_to_regions_cmd, freebayes_cmd, rm_regions_cmd])
def command(self):
return "picard -XX:ParallelGCThreads=1 CollectWgsMetrics " + \
required("I=", self.input) + \
required("R=", self.reference_sequence) + \
required("O=", self.output_metrics) + \
optional("MINIMUM_MAPPING_QUALITY=", self.minimum_mapping_quality) + \
optional("MINIMUM_BASE_QUALITY=", self.minimum_base_quality) + \
optional("COVERAGE_CAP=", self.coverage_cap)
def command(self):
required("input_files", self.input_files)
required("output_base", self.output)
required("dir_to_search", self.search_dir)
basefn = os.path.basename(self.output)
odir = os.path.dirname(self.output)
return "multiqc " + \
required("", self.search_dir) + \
required("-o ", odir) + \
optional("-n ", basefn) + \
optional("-k ", self.data_format) + \
optional("-i ", self.report_title) + \
" --data-dir --zip-data-dir -v -f"
def command(self):
return "picard -XX:ParallelGCThreads=1 -Xmx5g CollectGcBiasMetrics CHART=/dev/null" + \
required("I=", self.input) + \
required("O=", self.output_metrics) + \
required("S=", self.output_summary) + \
required("R=", self.reference_sequence) + \
optional("STOP_AFTER=", self.stop_after)
def command(self):
return "cat " + \
required(" ", self.input) + \
" | bgzip " + \
required(" > ", self.output) + \
" && tabix " + \
optional("-p ", self.filetype) + \
" {} ".format(self.output)
def command(self):
return "picard -XX:ParallelGCThreads=1 CollectHsMetrics " + \
required("I=", self.input) + \
required("R=", self.reference_sequence) + \
required("O=", self.output_metrics) + \
required("TI=", self.target_regions) + \
required("BI=", self.bait_regions) + \
optional("BAIT_SET_NAME=", self.bait_name) + \
repeat('METRIC_ACCUMULATION_LEVEL=', self.accumulation_level)
def command(self):
return "gatk-klevebring -T HeterozygoteConcordance " + \
required("-R ", self.reference_sequence) + \
required("-V ", self.input_vcf) + \
required("-I ", self.input_bam) + \
required("-sid ", self.normalid) + \
optional("-L ", self.target_regions) + \
conditional(self.filter_reads_with_N_cigar, "--filter_reads_with_N_cigar") + \
required("-o ", self.output)
def command(self):
if not self.reference and not self.targets_bed:
raise ValueError("Either reference or targets_bed must be supplied")
if self.reference and self.targets_bed:
raise ValueError("Supply either reference OR targets_bed")
tmpdir = "{}/cnvkit-{}".format(self.scratch, uuid.uuid4())
sample_prefix = stripsuffix(os.path.basename(self.input_bam), ".bam")
cnvkit_cmd = "cnvkit.py batch " + required("", self.input_bam) + \
optional("-r ", self.reference) + \
conditional(self.targets_bed, "--fasta " + str(self.fasta) + " --split ") + \
conditional(self.targets_bed, "-n") + \
optional("-t ", self.targets_bed) + \
required("-d ", tmpdir)
copy_cns_cmd = "cp {}/{}.cns ".format(tmpdir, sample_prefix) + required(" ", self.output_cns)
copy_cnr_cmd = "cp {}/{}.cnr ".format(tmpdir, sample_prefix) + required(" ", self.output_cnr)
rm_cmd = "rm -r {}".format(tmpdir)
return " && ".join([cnvkit_cmd, copy_cns_cmd, copy_cnr_cmd, rm_cmd])
def command(self):
# activating conda env
activate_cmd = "source activate purecn-env"
# running PureCN
running_cmd = "PureCN.R " + required("--out ", self.outdir) + \
required("--sampleid ", self.tumorid) + \
required("--segfile ", self.input_seg) + \
required("--vcf ", self.input_vcf) + \
required("--gcgene ", self.gcgene_file) + \
required("--genome ", self.genome) + \
optional("--minpurity ", self.minpurity) + \
optional("--hzdev ", self.hzdev) + \
optional("--segfilesdev ", self.seg_sdev) + \
conditional(self.postopt, "--postoptimize")
# deactivating the conda env
deactivate_cmd = "source deactivate"
return " && ".join([activate_cmd, running_cmd, deactivate_cmd])
def command(self):
# activating conda env
activate_cmd = "source activate purecn-env"
# running PureCN
running_cmd = "PureCN.R " + required("--out ", self.outdir) + \
required("--sampleid ", self.tumorid) + \
required("--segfile ", self.input_seg) + \
required("--tumor ", self.input_cnr) + \
required("--vcf ", self.input_vcf) + \
required("--genome ", self.genome) + \
optional("--funsegmentation ", self.funseg) + \
optional("--minpurity ", self.minpurity) + \
optional("--hzdev ", self.hzdev) + \
optional("--maxnonclonal ", self.maxnonclonal) + \
optional("--minaf ", self.minaf) + \
optional("--error ", self.error) + \
conditional(self.postopt, "--postoptimize")
# deactivating the conda env
deactivate_cmd = "conda deactivate"
# touching required output files
touch_cmd = "touch {} {} {} {}".format(self.out_csv, self.out_genes,
self.out_variants, self.output)
return " && ".join(
[activate_cmd, running_cmd, deactivate_cmd, touch_cmd])
def command(self):
activate_env_cmd = "source activate qdnaseqenv"
qdnaseq_cmd = "qdnaseq.R " + \
required("--bam ", self.input) + \
required("--output ", self.output) + \
optional("--background ", self.background)
deactivate_env_cmd = "source deactivate"
return "{} && {} && {} ".format(
activate_env_cmd,
qdnaseq_cmd,
deactivate_env_cmd,
)
def command(self):
return "target_coverage_histogram.py " + \
required("--targets ", self.input_bed) + \
required(" ", self.input_bam) + \
optional("--min_basequal ", self.min_basequal) + \
required("> ", self.output)
